From: "Perry E. Metzger" <perry@piermont.com>
Date: Wed, 29 Nov 1995 02:52:12 +0800
To: Jonathan Zamick <JonathanZ@consensus.com>
Subject: Re: The future will be easy to use
In-Reply-To: <v02120d02ace0fcab0df8@[157.22.240.13]>
Message-ID: <199511281801.NAA00386@jekyll.piermont.com>
MIME-Version: 1.0
Content-Type: text/plain



Jonathan Zamick writes:
> This discussion was based on a group of people getting together to create
> a new easy to use package for handling keys and such. The government is
> going to try to take a dominant stance, and mandate elements of it.

So we can ignore tem. Big deal. They have no laws with which to
enforce their desires.

> However, it is possible, even in an antagonistic relationship, to develop
> positive feedback.

Who cares? An hour spent talking to an idiot from Washington is better
spent writing good code unless there is a law pending in congress, in
which case you are probably better off paying someone who knows what
they are doing to do the talking for you.

> Returning to the original topic though, do we want to get a smaller list
> together to spec out some ideas for the project that was discussed? A
> simple, transparent, tool which would allow people to use strong encryption
> without having to think about it?

You mean, like IPSEC/Photuris? I'll be running IPSEC (but sadly not
Photuris, although I'll be trying to port Aggelos Keromytis' version
at some point) on my laptop at the IETF meeting in Dallas (provided
that I can buy a laptop in time.)

There are three things we are currently missing in the architecture,
IMHO.

1) We need a certificate system to replace X.509 and that plays nicely
   with distributed databases.
2) We need to implement the Eastlake/Kaufman method for embedding
   certificates in the DNS or something similar.
3) We need a good entity naming model.

Given all those being implemented, sometime soon I can see people
telnetting or ftping hither and thither without ever noticing or
caring that their sessions are completely encrypted.

We also have the following need:

4) A good MIME mailer (that looks like NeXT Mail or something like it)
   which has hooks for something MOSSlike that uses the same
   certificate infrastructure described in 1-3 above.
5) SHTTP capable browsers that also use 1-3 listed above.

.pm